Project 4: Project Summary/Abstract Bone marrow transplantation provided the first irrefutable evidence that cell therapy can mediate potent and long-lasting anti-cancer effects. Based upon these seminal observations, the field has worked diligently to understand and enhance cell mediated graft-versus-tumor (GVT) effects in the context of allogeneic and autologous stem cell transplantation. Despite these efforts, clinical benefit from GVT in B cell malignancies remains largely limited to patients with low burden disease who undergo dose intensive transplant conditioning following by infusion of autologous or allogeneic stem cell rescue. Since the first observation in 2011, it has become increasingly clear that T cells genetically engineered to express chimeric antigen receptors (CARs) can mediate potent and durable effects against B cell malignancies, establishing cell therapy for cancer as a viable therapeutic modality, even for patients with chemorefractory, high burden disease. Thus, emergence of CAR-T cell therapies is a natural extension of the efforts to harness cell based anti-tumor immunity that has long dominated research in the context of stem cell transplantation. Very simply, this Project seeks to improve short- and long-term benefit of CAR based therapies for B cell malignancies. Experience in B-ALL has identified two major mechanisms of CAR resistance, which appear to be largely mutually exclusive: tumor escape due to loss or diminished expression of the targeted antigen (which typically occurs in the presence of persistent CAR-T cells) and T cell failure (which typically occurs with continued expression of the targeted antigen). Aim 1 will test the hypothesis that antigen level is an important factor impacting response and relapse following CD19-CAR and CD22-CAR for DLBCL, and will test the efficacy of the first bispecific CAR to enter the clinic, with the goal of diminishing the risk of relapse due to antigen neg/lo variants in both B-ALL and DLBCL. Aim 2 tests the hypothesis that ?T cell exhaustion? is the major cause of relapse associated with T cell failure. We will seek to identify predictive biomarkers that can distinguish patients whose CAR T cells are predisposed to exhaustion, and undertake state-of-the-art single cell TCR/RNA sequencing to fate map persistent CAR T cells, as a first step toward a long-term goal of engineering grafts for ?exhaustion resistance?. Aim 3 begins to address a major practical challenge facing clinicians treating patients with CD19-CAR T cells for DLBCL, namely ?which patients should undergo post-CAR consolidation with autologous or allogeneic HSCT??. Building upon previous successes in this Program Project Grant in demonstrating the utility of circulating tumor DNA (ctDNA) in predicting clinical outcomes for B cell malignancies, we will test whether ctDNA can predict outcomes following CD19-CAR therapy for DLBCL, as a first step toward personalized stratification of post-CAR consolidation therapy.